Preparing for 0.0.1

Author: mpourmpoulis

README.md

@@ -2,7 +2,9 @@
 
 [![License](https://img.shields.io/badge/License-BSD%202--Clause-orange.svg)](https://opensource.org/licenses/BSD-2-Clause)
 
-![]( https://img.shields.io/static/v1?label=Caster&message=0.5.11&color=blue )
+![]( https://img.shields.io/static/v1?label=Caster&message=0.5.11&color=blue)
+
+![]( https://img.shields.io/static/v1?label=Version&message=0.0.1&color=green)
 
 PythonVoiceCodingPlugin is a Sublime Text 3 plugin meant to enhance user experience 
 when coding python 3 by voice. 
@@ -20,7 +22,7 @@ need to be taken to get there. What if you could instead simply describe (syntac
 ![](doc/gif/big3.gif)
 
 PythonVoiceCodingPlugin tries to enable you to do just that!
-To provide this functionality, it ships with bundles the implement a grammar, hopefully expressive enough for describing regions of interest, while running on the voice coding macro system side  and 
+To provide this functionality, it ships with bundles the implement a grammar, hopefully expressive enough for describing regions of interest, while running on the voice coding macro system side. These bundles
 cooperate with the core plugin, running on the editor side, arguably the more suitable of the two environments
 for analyzing source code and decoding the meaning of queries within the given context. 
 
@@ -59,7 +61,8 @@ Needless to say, while coding PythonVoiceCodingPlugin , PythonVoiceCodingPlugin
 
 The code is available on [github](https://github.com/mpourmpoulis/PythonVoiceCodingPlugin)
 along with its initial release (0.0.0)!
- 
+
+Update:  
 
 
 
@@ -140,9 +143,11 @@ So effectively we trigger the command our plugin provides in its top file and pa
 
 Of course this does not work across virtual machine barriers:) 
 
-to counter this issue the simplest idea, which involves only keystrokes, is to open with some key command sublime console and using a text command to invoke the PythonVoiceCodingPlugin command from the command line. But 
-This is often way to slow if there are a lot of parameters to pass and messes up with user-interface.
-so this is not implemented, even though  it does not need custom rpc.
+To counter this issue the simplest idea, which involves only keystrokes, is to open with some key command sublime console and using a text command to invoke the PythonVoiceCodingPlugin command from the command line. But 
+from my experience this is often way to slow if there are a lot of parameters to pass and messes up with user-interface, as it opens up the console.
+
+So this is not implemented, even though  it does not need custom rpc.
+
 
 For the time being I have bundles for 0.5.11 release of Caster that work as expected on Windows 10 64bit.
 

doc/SelectBigROI.md

@@ -1,6 +1,6 @@
 # Select Big Region of Interest Command
 
-These commands are responsible for selecting "big" regions of interest. By "big" we refer to things like "if condition" or "assignment right" in contrast to say the first entry of a dictionary.
+These commands are responsible for selecting "big" regions of interest. By "big" we refer to things like "if condition" or "assignment right" in contrast to say the first entry of a dictionary, though we will see later, by means of subindexes we can select [smaller stuff!](#SubIndexes)
 
 ![](./gif/big2.gif)
 
@@ -13,7 +13,7 @@ Choice("big_roi",{
 "else if condition" : "else if condition", 
 "while condition" : "while condition", 
 
-"if expression condition" : "if´´ expression condition", 
+"if expression condition" : "if expression condition", 
 "if expression body" : "if expression body", 
 "if expression":"if expression",
 
@@ -64,16 +64,8 @@ As you might expect, the plugin will try to find matches to big roi description
 
 ![](./gif/big1.gif)
 
-However,there are cases where you are interested in only a portion of this whole ROI. For instance there may be multiple targets in the left hand of an assignment and you want to select only one of them. Or your function may return multiple values. In those cases the 
+t is also important to note that with exception of the import statements, all other queries search only within the current function.
 
-```python 
-IntegerRefST("big_roi_sub_index",0,10),
-```
-comes in useful:
-
-![](./gif/big2.gif)
-
-it is also important to note that with exception of the import statements, all other queries search only within the current function.
 
 # Case two 
 
@@ -148,3 +140,41 @@ Once we have established which function we are to search, the command will then
  but will search inside that function!
 
 ![](./gif/big5.gif)
+
+# SubIndexes
+
+So far we have seen cases where we can select an entire region of interest.
+However,there are cases where you are interested in only a portion of this whole ROI. For instance there may be multiple targets in the left hand of an assignment and you want to select only one of them. Or your function may return multiple values. In those cases the 
+
+```python 
+IntegerRefST("big_roi_sub_index",0,10),
+```
+comes in useful:
+
+
+![](./gif/big2.gif)
+
+Okay , can we do something similar for other cases? like select only a portion of an if condition?
+
+![](./gif/big6.gif)
+
+As illustrated above, you need to pay attention to how the various conditions are bound together
+(or binds weaker causing it to be higher in the AST) in the can only select smaller conditions  from the outermost level!
+
+This feature existed ever since the initial release but was only documented on 0.0.1 . This release also expanded the feature from applying only to ast.BoolOp nodes to encompass ast.Compare nodes as well! in plain English:
+
+![](./gif/big7.gif)  
+
+furthermore, big_roi_sub_index can make our lives easier even in cases like the one below:
+
+![](./gif/big8.gif)  
+
+where we want to play with the indexes of a subscript!
+
+Finally, we clarify one more thing! What about relative vertical offsets when using above? We know that these abstract vertical keywords only count interesting lines, but what do we count as interesting here? To stay compatible with all of the above, we count all lines containing our desired big region of interest regardless of whether we can extract or not from them information with the sub index! As an example:
+
+![](./gif/big9.gif)  
+
+
+
+

doc/gif/big7.gif

@@ -14,6 +14,8 @@ def build_tree(code : str):
     return root,atok
 
 def get_source_region(atok, element):
+    if element is None:
+    	return None
     if isinstance(element,ast.AST):
         return atok.get_text_range(element)         
     else:

doc/gif/big8.gif

@@ -149,44 +149,50 @@ def after_bracket(t):
 
 
 def before_dot(t):
-	# return t[ - 1] is None  or not t[-1].string in ["from",".","import"]
+	return t[ - 1] is None  or not (
+		t[-1].string in ["from",".","import"] or 
+		finish_atom(t[-1])
+	)
 	return False
 
 def after_dot(t):
-	# return t[1] is None  or not (
-		# t[1].string in ["."] and 
-		# t[-1] is not None and
-		# t[-1].string in ["from",".","import"]
-	# )
+	return t[1] is None  or not (
+		start_atom(t[1]) or (
+			t[1].string in ["."] and 
+			t[-1] is not None and
+			t[-1].string in ["from",".","import"]
+		)
+	)
 	return False
 
 def handle_empty_compound(atok ,t,l,b,dummy):
 	n = neighbors(atok, t)
 	left,right = expand_to_line_or_statement(atok,t, l, b)
 	if token.DEDENT==left.type:
 		left = next_token(atok,left)
-	# print("empty compound ",left,right)
-	if t.string=="elif":
+	print("empty compound ",[left,right])
+	if t.string=="elif"  and left.string!="elif":
 		left = next_token(atok,left)
 	if left is t  and right.string == ":" :
 		rh = next_token(atok,right)
 		# print("rh is",[rh])
-		while rh and (rh.line.isspace() or rh.line == right.line):
+		while rh and (rh.line.isspace() or rh.start[0]==right.start[0]):
 			# print("rh is",[rh])
 			rh = next_token(atok, rh)
 		temporary = left.line
 		ls = temporary[:len(temporary) - len(temporary.lstrip())]
 		temporary = rh.line if rh else ""
 		rs = temporary[:len(temporary) - len(temporary.lstrip())]
-		# (print("\nstarting new variation\n",[t],"\n",ls,rs,"these are the indentations",len(ls),len(rs),[left,rh]))
+		(print("\nstarting new variation\n",[t],"\n",ls,rs,"these are the indentations",len(ls),len(rs),[left,rh]))
 		if len(ls)>=len(rs):
 			return True , right.endpos," pass ", False
 		else:
 			return (False,)
-	elif t.string=="if" and right is t:
-		return True , right.endpos, " " + dummy + " else " + dummy,True
-	else:
 
+	# these needs to be updated
+	elif t.string=="if"  and   right is t:
+			return True , right.endpos, " " + dummy + " else " + dummy,True
+	else:
 		return (False,)
 
 

doc/gif/big9.gif

@@ -84,9 +84,30 @@ def modified_information(x, information,index):
 					elif match_node(data,(ast.BoolOp)) :
 						if index<len(data.values):
 							return data.values[index]
+					elif match_node(data,(ast.Compare)) :
+						temporary = [data.left] + data.comparators
+						if index<len(temporary):
+							return temporary[index]
+
+					elif match_node(data,(ast.Subscript)):
+						data = data.slice
+						if match_node(data,(ast.Index)):
+							temporary = [data.value] 
+							if match_node(data.value,(ast.List,ast.Tuple,ast.Set)):
+								temporary  =  data.value.elts
+						elif match_node(data,(ast.Slice)):
+							temporary = [data.lower,data.upper, data.step]
+						elif match_node(data,(ast.ExtSlice)):
+							temporary = data.dims
+						temporary = [x  for x in temporary if x]
+						if index<len(temporary):
+							return temporary[index]
+
 					else:
 						return None
-				return (temporary[0],temporary[1], lambda x: modified_information(x,temporary[2],index-1))
+				y  = lambda x: temporary[2](x)
+				y.secondary  = lambda x: modified_information(x,temporary[2],index-1)
+				return (temporary[0],temporary[1],y)
 
 
 	def case_one(self,view_information,query_description, extra = {}):
@@ -95,6 +116,7 @@ def case_one(self,view_information,query_description, extra = {}):
 		###############################################################	
 		build, selection, origin, definition_node = self.preliminary(view_information, query_description,extra)
 		targets, exclusions, information  =  self.decode(query_description)
+		information = getattr(information,"secondary",information)
 		candidates = tiebreak_on_lca(definition_node,origin,find_all_nodes(definition_node, targets, exclusions))
 		candidates = [information(x)  for x in candidates if information(x)]
 		result, alternatives = obtain_result(None, candidates)
@@ -123,6 +145,7 @@ def case_two(self,view_information,query_description, extra = {}):
 				information_nodes = find_matching(definition_node,temporary_information),
 				**additional_parameters
 		)
+		information = getattr(information,"secondary",information)
 		result = information(result) if result else None
 		alternatives  =[ information(x)  for x in alternatives] if alternatives else []
 		return  self._backward_result(result, alternatives,build)
@@ -149,6 +172,7 @@ def case_three(self,view_information,query_description, extra = {}):
 				alternatives  = tiebreak_on_lca(new_definition_node,result,find_all_nodes(new_definition_node,targets , exclusions))
 
 		result, alternatives = obtain_result(result, alternatives)
+		information = getattr(information,"secondary",information)
 		result = information(result) if result else None
 		alternatives  = [information(x)  for x in alternatives] if alternatives else []
 		return  self._backward_result(result, alternatives,build)
@@ -170,6 +194,7 @@ def case_four(self,view_information,query_description, extra = {}):
 
 		t = decode_abstract_vertical(root,atok,targets,row, ndir + bonus,direction,True,temporary_information)
 		if query_description["adjective"]=="None":
+			information = getattr(information,"secondary",information)
 			candidates = tiebreak_on_lca(root,definition_node,find_all_nodes(t, targets, exclusions))
 			candidates = [information(x)  for x in candidates if information(x)]
 			result, alternatives = obtain_result(None, candidates)
@@ -184,6 +209,7 @@ def case_four(self,view_information,query_description, extra = {}):
 				information_nodes = find_matching(t,lambda x: information(x) if match_node(x,targets,exclusions) else None),
 				**additional_parameters
 			)
+			information = getattr(information,"secondary",information)
 			result = information(result) if result else None
 			alternatives  =[ information(x)  for x in alternatives] if alternatives else []
 			return  self._backward_result(result, alternatives,build)

library/__init__.py

@@ -1,10 +1,10 @@
 def obtain_result(result, alternatives):
 	if result:
-		return result, alternatives if alternatives is not None else []
+		return result, [x  for x in  alternatives if x is not result] if alternatives is not None else []
 	else:
 		if alternatives is None or len( alternatives )==0:
 			return None,None
 		else:
-			return  alternatives[0], alternatives[1:]
+			return  alternatives[0], [x  for x in alternatives[1:] if x is not alternatives[0]]
 
 # bug if result is in alternatives